Technical SEO

A lot of SEO advice is about content and backlinks — write more, get more sites linking to you. Both matter, eventually. But neither one fixes a page that a search crawler can't properly read, or a site so slow that Google marks it down before a single word of the content is even evaluated. That foundation is what technical SEO covers, and it's usually invisible until it's broken.

Why the technical foundation comes first

Google's ranking system uses Core Web Vitals — load speed, interactivity, visual stability — as a direct ranking factor, and separately needs to be able to crawl and render a page correctly to understand what it's about at all. A site built as a client-side JavaScript app that assembles its content in the visitor's browser can look completely normal to a human, while a crawler sees a mostly empty page, because the actual content hasn't rendered yet by the time it's evaluated. No amount of well-written content fixes that — the content was never actually visible to the thing doing the ranking.

This is the direct, practical reason Next.js and server-side rendering matter for SEO specifically: pages get pre-rendered as real HTML before a crawler ever asks for them, so there's no gap between what a visitor sees and what Google sees.

What's actually included

Rendering strategy. Making sure every page that needs to rank is rendered as real, crawlable HTML — not assembled client-side after the fact — and choosing the right strategy (static generation, server rendering, or incremental regeneration) per page based on how often its content actually changes.

Core Web Vitals. Image optimization, code splitting so pages don't ship JavaScript they don't need yet, and layout stability so nothing shifts around as a page finishes loading — the concrete, measurable levers behind the score Google actually uses.

Structured data. Schema.org markup — FAQPage, BreadcrumbList, Article, LocalBusiness, and others depending on the content — that tells search engines explicitly what a page contains instead of leaving them to infer it. This site uses exactly that: every service and landing page ships FAQPage and BreadcrumbList schema, which is part of why FAQ answers here are eligible to show up directly in search results instead of just a blue link.

Sitemaps, robots rules, and canonical tags, kept in sync with what actually exists on the site instead of drifting out of date as pages get added or removed — a stale sitemap pointing at deleted pages, or a missing canonical tag on duplicate content, quietly wastes crawl budget and confuses ranking signals.

Multilingual SEO, where relevant: correct hreflang tags so Google shows a French visitor the French version of a page and a German visitor the German one, instead of both competing against each other for the same ranking or, worse, being treated as duplicate content.

Programmatic SEO pages for long-tail search terms — pages generated systematically by combining real dimensions of a business (category × location, service × city) rather than one generic page trying to rank for everything at once.

Search intent, matched at the structural level

Different searches expect different kinds of pages, and a lot of avoidable underperformance comes from a mismatch between the two — a single generic "services" page trying to rank for a search that's actually looking for a specific answer, or a blog post trying to rank for a search that's actually looking for a page to take action on (book, buy, contact). Matching structure to intent means recognizing which searches want a comparison, which want a direct transactional page, and which want an explanation — and building the page type that actually fits, rather than funneling every search term toward the same homepage and hoping relevance sorts itself out. This is a structural decision made before a single word of content is written, and it's a large part of why the programmatic pages described below are built as category-and-location combinations rather than one page per category — each combination matches a specific, real search intent instead of approximating several at once.

A real example: 40+ pages, one new domain, no ad budget

HorecaJob.md is a trilingual job platform for the hospitality industry in Moldova, and its organic acquisition strategy is a direct example of this approach rather than a hypothetical one. Instead of a single "jobs" page, the platform generates landing pages by combining job category (waiter, cook, barista) with city — over 40 of them at launch. For a brand-new platform with no existing audience and no marketing spend, that's not a nice-to-have; it's close to the only realistic way to get found at all, since there's no other traffic source to lean on while authority builds. The trilingual routing (Romanian, Russian, English) was designed into the URL and hreflang structure from the start rather than retrofitted, which avoids the duplicate-content and cannibalization problems a bolted-on translation layer tends to create.

On-page structure: still technical, not copywriting

Between "technical infrastructure" and "written content" sits a layer that's still structural rather than creative: making sure each page's title tag and meta description actually reflect what someone searching would type, that heading hierarchy (one H1, logically nested H2s and H3s beneath it) matches the actual structure of the content instead of being chosen for visual size, and that internal links between related pages exist and use descriptive text instead of generic "click here" links that tell a crawler nothing about what the linked page contains. None of that is copywriting in the sense of persuasive sales language — it's making sure the structure a page already has is legible to both a visitor scanning it and a crawler indexing it. A page can have genuinely good content and still underperform in search simply because none of this scaffolding around it was done.

Measuring what actually happens after launch

Technical SEO work is only verifiable if it's measured. That means analytics wired up correctly from day one — page views, but more importantly the events that indicate real intent: a contact form submitted, a pricing page viewed, a phone number clicked — and Google Search Console connected so indexing status and search query data come from Google directly instead of being inferred. For sites also running region-specific tools (Yandex Metrika alongside Google Analytics is common for markets where both search engines matter), both get set up the same way: loaded early enough to catch a visitor who leaves within seconds, not just the ones who stay long enough for a slower-loading tracker to fire. Without that baseline in place before launch, there's no reliable way to tell whether a technical change actually moved anything, or whether a ranking change was the fix or just noise.

Migrations: where SEO most commonly gets quietly broken

Rebuilding or moving a site is one of the highest-risk moments for search rankings, and it's usually not the new site's fault when rankings drop afterward — it's a missed redirect. Every URL that existed before a migration needs an explicit destination after it; a page that returns a 404 instead of a 301 redirect doesn't just lose that one page's ranking, it loses whatever authority was pointing at it from other sites and pages. The same applies to a domain-level change — moving DNS providers, switching nameservers, or changing where a site is hosted — where the technical delegation has to be verified end to end (registry records, not just what a hosting panel's dashboard claims) before assuming a domain actually resolves the way it's supposed to.

Every migration handled here starts with a full map of existing URLs, redirects put in place before the old version comes down, and a verification pass on the new site's indexability before calling it done.

Local SEO, for businesses that need it

For a business that serves a specific city or region — a restaurant, a clinic, a contractor — technical SEO overlaps with a narrower, very concrete set of requirements: a correctly configured Google Business Profile matching the site's name, address, and phone number exactly, LocalBusiness structured data on the site itself so search engines can connect the two, and location-specific pages where the business genuinely serves multiple distinct areas rather than one page vaguely claiming to serve "the whole region." This is a smaller, more mechanical version of the programmatic-page approach described above — instead of category × city across a whole marketplace, it's usually a handful of well-structured location pages for a single business, each one specific enough to actually rank for its own area instead of all of them diluting each other by targeting the same broad terms.

What this isn't

This isn't a full-service SEO agency running monthly content calendars, backlink outreach campaigns, or open-ended keyword strategy retainers — that's a different, larger kind of engagement, and one this studio doesn't offer as an ongoing service. What's on offer is narrower and more technical: get the foundation right, so that content, when it's written, and backlinks, when they're earned, actually have something crawlable and fast to point at. A brilliant blog post sitting on a page that takes six seconds to load and doesn't render its content for crawlers is fighting itself before it even competes with anyone else.

---

Have a site that isn't showing up in search the way it should, or a migration coming up that makes you nervous about losing rankings? Get in touch — a first technical read on what's actually happening is free.